Self-organized Control of Manned and Unmanned Vehicles in Space Colonies PĂŠter MolnĂĄr Kameswara R. Namaduri Lance M. Kaplan Center for Theoretical Studies of Physical Systems Clark Atlanta University
Outline • •
• •
Motivation Concept of distributed assignment problem solver Implementation and application Status & Goals
Motivation • Space Colonies will host a multitude of mobile entities vehicles • space ships: usually manned • service vehicles: pickup cargo and passengers, transport fuel and other supplies • maintenance and repair robots, etc. • Navigation challenges • collision avoidance (basic navigation) • path planning: short travel distances, avoid congestion • Assignment and scheduling problems: gate assignment, dispatcher for service robots.
Space Port
2D Assignment Problem
New Concept • Central Control • vehicles dispatched to individual destinations, • rigid schedule, • dependency on control center.
• Self-organization • vehicles select destinations, • schedule emerges, • adaptable to changes & use of spare vehicles, • inc. complexity and number of vehicles.
Distributed Assignment Problem Solver •
•
•
Robots create initial preferences (priority list), preferences will be exchanged via wireless communication. Robots update their preference based on the received information of other units. In the time limit the system always converges to a feasible solution.
Features •
•
•
•
CSE's can solve NP-hard problems: multidimensional assignments. Distributed algorithm, system can be partitioned. Resistant to malfunctions of single units, ability to cover up with spare units. Insensitive to communication problems.
Implementation •
•
•
•
Vehicles require positioning system and wireless communication. Unmanned vehicles need navigation system, that ties in the assignment problem solver. Pilots receive instruction, similar to messages from a control tower. Development: Integrated device & communication protocol/language.
Applications • • • •
Construction of space colonies. Operation: space port. Maintenance & first alert response. Use of versatile self-configurable and combinatorial robots.
Goals & Status • Develop simulation program. ½ ' • Identify scenario(s) & design simulation(s). • Test system for various conditions: • failure resistance (vehicles and communication) • performance w/ respect to complexity • number of vehicles